Learning Curve Calculator (Wright's Law, Time per Unit)
How the average time per unit falls as repetitive work is repeated, from the first-unit time and the learning rate, with the total time and the number of units to reach a target.
Results
The dashed line marks the value you entered.
What this tool does
Wright's learning curve, first described in the 1930s from aircraft production, says that each time cumulative output doubles, the average time per unit falls to a fixed percentage of what it was: an 80 % curve means 100 hours for the first unit, 80 for the average of two, 64 for the average of four. It describes repetitive manual work well at first and flattens as the easy gains run out. The learning rate is something you measure from your own earlier batches, not a constant of nature, and machine-paced work learns far less.
Formula
average time of the first n units = T₁ × nᵇ, b = ln(learning rate) ÷ ln 2 · total = n × average · (Wright's cumulative-average model)
Variables
| Symbol | Meaning | Unit |
|---|---|---|
t1 | Time to make the first unit | h |
lr | Learning rate (average time falls to this % each time output doubles) | % |
n | Number of units made | — |
tt | Average time per unit you want to reach | h |
AV | Average time per unit over those units | h |
TT | Total time for those units | h |
LU | Time for the last unit on its own | h |
NU | Units to make to reach the target average | — |
B | Learning exponent b | — |
Worked example
- Time to make the first unit100 h
- Learning rate (average time falls to this % each time output doubles)80 %
- Number of units made8
- Average time per unit you want to reach40 h
- Average time per unit over those units51.20 h
- Total time for those units409.6 h
- Time for the last unit on its own35.46 h
- Units to make to reach the target average17
- Learning exponent b-0.3219
Limitations
- The result is an estimate based only on the values you type. Real situations often include factors this calculator does not know about.
- The default values are typical reference figures, not measurements of your situation. Replace them with your own data whenever you have it.